One consistent and disturbing finding in the suicide literature compels this investigation. Most older adults who take their own lives have an affective disorder, but the potential risk for suicide was unrecognized by family members and health professionals. The central thrust of our symptom detection model is that specific factors deter the detection, diagnosis, and treatment of depression in older adults. ROl funding is sought to test aspects of this model in a demographically and clinically heterogeneous sample of 1000 primary care patients 65 years of age and older. Beginning in August 2001, these patients were recruited into a naturalistic study, "Depression Outcome in Primary Care Elderly" (DPC) (NIMH ROl MH61429-O1A1, J. Lyness, M.D., P.I.). For this proposed ROl, we will collect data from informants who are members of the social networks of participants in the DPC study in order to examine informant detection of depressive disorders and symptoms of depression and anxiety. We will strive to recruit one informant for each subject in that study, and plan to collect data from more than 628 informants. We will ask the informants questions about themselves and the DPC participants' psychiatric symptoms. Specifically, we will examine informant detection of depression as a function of: the severity and history of patients' psychiatric disorders (Aim 1), patients' psychosocial (personality and social support) characteristics and physical health parameters (Aim 2), patients' demographic characteristics (Aim 3), and informants' psychiatric history and self-reported health (Aim 4). We will explore the role of informants' personality traits and attitudes toward mental illness, and differences between African-Americans and Whites will be explored. We hypothesize the detection will be poorer for: disorders that are less severe and first episode (Aim1); patients with certain personality traits (e.g., low extraversion; low openness to experience) or poor physical health (Aim 2); and men and unmarried participants (Aim 3). Detection will also be poorer when informants have no prior history of depression or are in poor physical health (Aim 4). For Aim 2, the mediating effect of social support will be examined. Confirmation of these hypotheses will indicate needed revisions in current approaches to late-life depression and suicide. Findings will help guide the development of screening instruments, educational and clinical interventions, and surveillance strategies to lessen the public health impact of unrecognized and untreated depression.